Impact of Interference on OFDM based Radars

Abstract

The goal of this work is to identify the possible degradations on a Wi-Fi based Passive Radar in the presence of an interferer. We assume that the signal-of-opportunity and the interference are 802.11ax compliant. The mathematical model derived for the interference shows that in a synchronized case, the interference may yield ghost targets. When a more occasional interference scenario is considered, the range/Doppler Map accuracy decreases significantly. Furthermore, numerical results are provided to quantify certain effects of OFDM radar interference on range/Doppler Maps.

A multi‐antenna super‐resolution passive Wi‐Fi radar algorithm: Combined model order selection and parameter estimation

“…In order to save space, and focus on more important parts of the system model, we assume that the received OFDM signals already went through the following process: i) correlation is applied to the Short Training Fields to find the starting time of the received OFDM frames; ii) Carrier Frequency Offset is estimated from the High‐Efficiency Long Training Fields and compensated accordingly; iii) Cyclic Prefix samples are dropped, and the Fast Fourier Transform (FFT) of each OFDM symbol is computed; iv) through zero‐forcing, the CTF is estimated [27, 28], which can be written as with q = 0… Q − 1, n = 0… N − 1, m = 0… M − 1. $$$\begin{array}{r}\hfill \widehat{H}[q,n,m]=\sum\limits _{p=1}^{P}\left({\alpha }_{p}\,\omega \left(\frac{q{\tau }_{p}}{QT}\right)\,\omega \left(n\mathrm{sin}\left({\theta }_{p}\right)\right)\,\omega \left({f}_{p}m{T}_{f}\right)\right.\\ \hfill +\left.z[q,n,m]\right)\end{array}$$$ …”

“…In order to save space, and focus on more important parts of the system model, we assume that the received OFDM signals already went through the following process: i) correlation is applied to the Short Training Fields to find the starting time of the received OFDM frames; ii) Carrier Frequency Offset is estimated from the High‐Efficiency Long Training Fields and compensated accordingly; iii) Cyclic Prefix samples are dropped, and the Fast Fourier Transform (FFT) of each OFDM symbol is computed; iv) through zero‐forcing, the CTF is estimated [27, 28], which can be written as with q = 0… Q − 1, n = 0… N − 1, m = 0… M − 1. …”

High Resolution 802.11ax-based Passive Radar for Human Movement Monitoring

“…Then, each estimated CIR, with index m, is stacked on a matrix, also known as range/slow-time map. For a fixed range d i.e., tap index, the slow-time response can be written as [7] h…”

“…The set of frequencies and the number of signals are estimated by the 2-step ESPRIT. Corresponding Vandermonde matrix, Ã, is constructed from the estimated frequencies, as in (7), and the amplitudes, α, are estimated with the LS method through (5). Then, a confidence analysis is performed to determine the accuracy of each estimated frequency.…”